R - Latent Growth Models Lecture

[@amanrastogi5184]

Thank You Ma'am.

[@simongetch]

Thanks for your explicit explanation. My question here is do all types of longitudinal data be used directly with out any transformation? what are the assumption behind the nature of data distribution?

[@StatisticsofDOOM]

Most SEM procedures assume normal-ish distributions - is that what you mean? Or the type of growth?

[@simongetch]

@@StatisticsofDOOMThanks for your prompt response. ----The type of growth curve was my question.

[@user-wangchuan]

thanks a lot

[@zihengzhang2113]

Thank you so much for this lecture on latent growth model. I just have one question not quite clear in the model assumption part. You mentioned that the time spacing is the same across people, does it mean that all people needs to be measured in the same time scale (days vs. weeks), or all people need to be measured in the same time space ( day 1, day 2)? for example, if i measured some kind of exercise score of people, do all people need to be measured in day 1, day2 ………; or some people can be measured in day 1,2,5,8,10, and others are measured in day 1,3,4,5,8,9. Does this kind of measurement violate the assumption of this model. Thanks

[@StatisticsofDOOM]

All people within the same space - you can space the time between measurements differently and account for it within the slope coefficients. I would say that having different spacing between individuals would mean you'd need to model each group differently.

[@a45701]

Actually I have the same question, say you have the same 10 measuring days, but for some people you have missing data points, say one subject missed day 5. Can Lavaan handle this situation within the growth curve model without dropping people with incomplete records? Thanks, for the nice practical presentation.

[@tanyachichekian9900]

Would it be possible to produce a video on how to conduct a latent transition analysis?

[@StatisticsofDOOM]

I'm not aware of that type of analysis, but I can add it to look at.